PROJECT 1: SUMMARY/ABSTRACT Approximately 5% of patients of with Alzheimer's Disease (AD) develop symptoms before age 65 in the absence of a known autosomal dominant mutation. Patients with non-familial early-onset AD (EO-AD) show greater deficits in executive function, attention, language and visuospatial abilities and relatively spared episodic memory compared to patients with late-onset AD (LO-AD). Non-amnestic and hippocampal-sparing AD phenotypes are far more common in patients with EO disease. Accurately diagnosing AD in EO patients is challenging due to the atypical clinical presentations and overlap with non-AD dementia, and misdiagnosis rates are high even at expert centers. Molecular and neurodegenerative biomarkers are likely to facilitate early and accurate diagnosis, but few studies have addressed their performance in patients with EO disease - results from LO-AD studies may not generalize to EO populations because of differences in degenerative patterns and reference ranges. Furthermore, patients with EO-AD may harbor unrecognized susceptibility factors that lead to disease onset at such a young age. While the apolipoprotein E ?4 allele is strongly correlated with young age-of-onset, it is present in only ~50% of EO patients, suggesting that additional mechanisms of vulnerability have yet to be discovered. Leveraging on the strength of the UCSF ADRC in recruiting and characterizing patients with early-onset AD, this study will apply detailed clinical phenotyping, multi-modal neuroimaging and novel genetic approaches to optimize early-stage diagnosis and to elucidate mechanisms of vulnerability in EO-AD. We will evaluate 100 mildly impaired (CDR 0.5-1) early-onset (estimated age of onset ?65) patients recruited from the Clinical core. All patients will meet NIA-AA criteria for MCI or probable AD and demonstrate evidence of AD pathology based on a positive amyloid (florbetapir) PET scan. Comparative data from 100 matched normal controls (NC) and 75 non-AD dementia patients will be acquired via the Clinical and Imaging cores. Aim 1 will test the diagnostic performance of (1) CSF and (2) hippocampal versus cortically-based MRI biomarkers in distinguishing EO-AD from NC and non-AD dementia. Aim 2 will build on preliminary data suggesting that ApoE4 modifies the phenotype of EO-AD, testing the hypothesis that ApoE4 carriers will show greater memory impairment and hippocampal atrophy and lower amyloid compared to E4 non-carriers. A hypothesis-generating Aim 3 will apply novel genetic tools to study vulnerability factors in EO-AD, hypothesizing that: (1) ApoE4-positive and E4-negative patients will show differential gene expression patterns in peripheral blood, reflecting involvement of distinct biological pathways; and (2) by screening EO-AD patients with a gene chip that includes ~250,000 rare, protein-altering genetic markers, we will identify novel risk variants in genes implicated in AD pathogenic pathways. Overall, this project will facilitate the early and accurate diagnosis of EO-AD, and will further our understanding of susceptibility factors associated with pre-senile disease onset.